The present invention relates generally to an illumination optical system, and more particularly to an illumination optical system for illuminating a reticle (or a mask) which forms a pattern, in an exposure apparatus used in a photolithography process for fabricating semiconductor devices, liquid crystal display devices, image pick-up devices (CCD, and the like), thin-film magnetic heads, and the like.
The photolithography technology for manufacturing fine semiconductor devices, such as LSIs and very large scale integrations, has conventionally employed a reduction projection exposure apparatus that uses a projection optical system to project and transfer a circuit pattern formed on a reticle onto a wafer, etc. As the recent improved packaging density of the semiconductor devices requires finer patterns, the exposure apparatus needs to improve the resolution (to correspond to the fine processing).
The improved resolution of the exposure apparatus generally requires optimizations of both the numerical aperture (“NA”) of the projection optical system and the NA of the illumination optical system. Concretely, the illumination optical system optimizes the resolution and the contrast for a certain circuit pattern by adjusting a value of the coherence factor σ that corresponds to a ratio between the NA of the projection optical system and the NA of the illumination optical system. For example, an illumination optical system proposed in Japanese Laid-Open Patent Application No. 2002-217085 (corresponding to published United States Application No. 2002/109108) typically has a σ consecutively variable optical system that can continuously change a σ value.
FIG. 20 is a simplified optical-path diagram of a σ variable optical system 1000. The σ variable optical system 1000 has, in order from an exit side of a columnar glass HCD that has section shape of hexagon, an aperture stop 1010, a parallel plate 1020, a first lens unit 1100 that has a convex lenses 1110 and 1120, a second lens unit 1200 that has a concave lens 1210, a third lens unit 1300 that has convex lenses 1310 and 1320, and a fourth lens unit 1400 that has a concave lens 1410 and a convex lens 1420.
The σ variable optical system 1000 can continuously change a size of an irradiated area (illumination area) L or a value of σ by moving the concave lens 1210 in the second lens unit 1200 in the direction of arrow A along the optical axis, and by moving the convex lenses 1310 and 1320 as one member in the third lens unit 1300 in the direction of arrow B along the optical axis. FIG. 20A shows the minimum σ state that minimizes the irradiated area L, FIG. 20B shows the maximum σ state that maximizes the irradiated area L, and FIG. 20C shows the intermediate state in which the irradiated area L is between the minimum σ state and the maximum σ state.
The σ variable optical system 1000 of the configuration shown in FIG. 20 forms a re-condensing position ACP between the second lens unit 1200 and the third lens unit 1300 that are movable units, for the light emitted from a secondary light source position TLP that is a condensing position of the light irradiated from a light source. The re-condensing position ACP becomes the closest to the lens in the second lens unit 1200 or the third lens unit 1300, in the maximum σ state shown in FIG. 20C.
The high light energy density at the re-condensing position causes deteriorations in the internal transmittance of an optical material that composes the lens, and anti-reflection coatings applied onto the lens surfaces. In other words, the re-condensing position located near the lens would damage the lens, deteriorate its light intensity, and decrease the throughput of the exposure apparatus.
The illumination optical system of FIG. 20 arranges the re-condensing position between the movable units, as the above-mentioned, and has restrictions that it cannot expand a zooming range at a side of the high magnifying power of the σ variable optical system or a large σ side so as to prevent damages of the lens. Therefore, the illumination optical system may not possibly obtain the optimized σ or achieve the desired resolution for some circuit patterns.